Extensible automatic computing service platform and construction method for same

Abstract

The invention discloses an extensible automatic computing service platform and a construction method for the same. The extensible automatic computing service platform comprises a material resource layer, a virtual resource layer, a platform management layer and a business service layer. According to the extensible automatic computing service platform and the construction method for the same, the computation of a computing task and the automatic creation of a computing tool running environment are unified, so that a user can conveniently and rapidly construct a computing tool running environment suitable for the computing task; heterogeneous resources are managed in a unified manner, and are abstracted and unified by virtue of a virtualization technology, so that unified resources are provided for the user to call; good external interfaces are packaged, and task-related functional interfaces are packaged, so that difficulty in the use of the user is greatly lowered, and the programming of the user for resource calling and automatic task computation is facilitated; a service-oriented hierarchical architecture is adopted for the platform, so that high extensibility is achieved, and computing resources can be conveniently extended.

Description

technical field [0001] The invention belongs to the field of computer science and technology, and in particular relates to an expandable automatic computing service platform and a construction method thereof. Background technique [0002] At present, with the increasing complexity of computing problems, the amount of computing data shows a geometric growth, which brings a series of computing-intensive, data-intensive, and communication-intensive computing problems in spatial information science. Using parallel computing technology to solve computing problems is one of the most important ways. Parallel computing has the advantages of simple principle, strong computing power, flexibility and convenience, but each parallel computing tool also has defects such as low resource utilization and poor scalability, especially for problems such as heterogeneous computing environments and unequal computing resource requirements. Faced with these problems, most of the existing solutions...

AI Technical Summary

Problems solved by technology

[0005] 1. Eucalyptus lacks API layer design, computing, storage, and network services are not separated, inconvenient to use, insufficient support for plug-ins, poor scalability, and poor support for users

[0006] 2. CloudStack does not adopt a service-oriented architecture design, nor does it independently design computing, storage, and networks. The horizontal expansion process of the platform is complicated and error-prone

[0007] 3. IBM's SKC is not an open source product, and the implementation is more complicated

[0008] 4. The installation of OpenStack is complicated, the API interface is slightly confusing, and the new version often changes greatly, making it difficult to upgrade and maintain

[0009] 5. All cloud computing platforms do not support automatic configuration and operation of computing environments and computing tasks, which imposes a heavy burden on computing service users

Images